Once colorectal cancer metastasizes, it becomes a lethal disease with a 5-year survival rate of approximately 10%. Effective therapeutics that can specifically target metastatic colorectal tumor cells are sorely needed. Delivery of nucleic-acids (e.g. genes or RNAi) to combat cancer is a highly promising therapeutic approach; unfortunately, targeted delivery of gene vectors to tumor cells has been largely difficult to achieve. Most vector targeting approaches to date have relied on cell surface receptors overexpressed on some subpopulation of target cancer cells. Unfortunately, there is no unique cell surface biomarker that specifically identifies all cells in a tumor. To overcome this limitatin, we propose to develop protease-activatable viruses (PAVs) that use extracellular proteases overexpressed in metastatic colorectal tumor microenvironments as the biomarkers to achieve targeted delivery. Specifically, matrilysin (also known as matrix metalloproteinase 7, MMP7) has been shown to be overexpressed in colorectal cancer. High levels of MMP7 in the tumor microenvironment will activate the PAVs in a localized manner and enable the vectors to bind cellular receptors that are broadly expressed, including on colorectal cancer cells, and mediate efficient gene delivery. Our PAV technology is based on the clinically promising adeno-associated virus (AAV), which has recently been approved as the first human gene therapy product in Europe. We have key pilot data demonstrating we have created MMP7-sensing PAVs that dramatically increase their gene delivery efficiency once exposed to the protease. Moreover, in an orthotopic cancer model, a PAV prototype is able to significantly increase transgene delivery and expression in tumors. In aim 1, we will synthesize and characterize a panel of MMP7-sensing PAVs. Our design process will harness both rational and combinatorial approaches in order to expedite achievement of the design solution. In aim 2, we will test the gene delivery performance of PAVs in vitro on colorectal cancer cells, and mechanistic studies will be done to probe the interaction of PAVs with the cells. Finally, we will test the PAVs in an orthotopic model of metastatic colorectal cancer in order to determine their in vivo specificity and therapeutic efficacy. If successful, this project will generate protease-responsive AAV vectors that may become viable therapeutic options for metastatic colorectal cancer.